microorganisms

Article Land-Use Type Drives Soil Population Structures of the Entomopathogenic Fungal Genus

María Fernández-Bravo 1 , Florian Gschwend 1 , Johanna Mayerhofer 1 , Anna Hug 2, Franco Widmer 1 and Jürg Enkerli 1,*

1 Molecular Ecology, Agroscope, CH-8046 Zürich, Switzerland; [email protected] (M.F.-B.); fl[email protected] (F.G.); [email protected] (J.M.); [email protected] (F.W.) 2 Swiss Soil Monitoring Network (NABO), Agroscope, CH-8046 Zürich, Switzerland; [email protected] * Correspondence: [email protected]

Abstract: Species of the fungal genus Metarhizium are globally distributed pathogens of , and a number of biological control products based on these fungi have been commercialized to control a variety of pest arthropods. In this study, we investigate the abundance and population structure of Metarhizium spp. in three land-use types—arable land, grassland, and forest—to provide detailed information on habitat selection and the factors that drive the occurrence and abundance of Metarhizium spp. in soil. At 10 sites of each land-use type, which are all part of the Swiss national soil-monitoring network (NABO), Metarhizium spp. were present at 8, 10, and 4 sites, respectively.   On average, Metarhizium spp. were most abundant in grassland, followed by forest and then arable land; 349 Metarhizium isolates were collected from the 30 sites, and sequence analyses of the Citation: Fernández-Bravo, M.; nuclear translation elongation factor 1α gene, as well as microsatellite-based genotyping, revealed Gschwend, F.; Mayerhofer, J.; Hug, A.; the presence of 13 Metarhizium brunneum, 6 Metarhizium robertsii, and 3 Metarhizium guizhouense Widmer, F.; Enkerli, J. Land-Use Type Drives Soil Population Structures of multilocus genotypes (MLGs). With 259 isolates, M. brunneum was the most abundant species, and the Entomopathogenic Fungal Genus significant differences were detected in population structures between forested and unforested sites. Metarhizium. Microorganisms 2021, 9, Among 15 environmental factors assessed, C:N ratio, basal respiration, total carbon, organic carbon, 1380. https://doi.org/10.3390/ and bulk density significantly explained the variation among the M. brunneum populations. The microorganisms9071380 information gained in this study will support the selection of best-adapted isolates as biological control agents and will provide additional criteria for the adaptation or development of new pest Academic Editor: Victor Glupov control strategies.

Received: 28 May 2021 Keywords: M. brunneum; M. robertsii; M. guizhouense; microsatellite; SSR; EF-1alpha; abiotic factors; Accepted: 22 June 2021 arable land; grassland; forest; biological control Published: 25 June 2021

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in 1. Introduction published maps and institutional affil- iations. The genus Metarhizium Sorok¯ın (: ) includes more than 30 described species (including both asexual and sexual states) that are pathogenic to arthropods, particularly and some arachnids [1,2]. The wide host spectrum of Metarhizium spp. includes many important crop pests such as Helicoverpa armigera (Hübner, 1805), Diabrotica virgifera virgifera (LeConte, 1868), Agriotes spp. (Eschscholtz, 1829), and Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. also termites, cockroaches, and even disease-transmitting insects such as tse-tse flies [2–7]. This article is an open access article Based on the pathogenicity of this , a number of biological control agents have been distributed under the terms and developed as commercial mycoinsecticides and mycoacaricides to control pest arthropods conditions of the Creative Commons worldwide (see [8] and https://www.eppo.int; https://www.epa.gov). Attribution (CC BY) license (https:// Soil is considered the main habitat and reservoir of Metarhizium spp., and they have creativecommons.org/licenses/by/ been isolated from very different ecosystems, from tropical to temperate, semiarid, and 4.0/). artic areas [5,9]. Besides being entomopathogenic, Metarhizium spp. exhibit a pronounced

Microorganisms 2021, 9, 1380. https://doi.org/10.3390/microorganisms9071380 https://www.mdpi.com/journal/microorganisms Microorganisms 2021, 9, 1380 2 of 19

saprophytic lifestyle, and, for some isolates and/or species, symbiotic interactions with plants as rhizosphere colonizers and/or endophytes have been reported [9,10]. Studies have demonstrated, for instance, that isolates of M. robertsii J.F. Bisch., S.A. Rehner & Humber, M. brunneum Petch, and M. guizhouense Q.T. Chen & H.L. Guo can provide - derived nitrogen to plants [11,12], increase plant growth and productivity [13,14], as well as protect plants from environmental stresses, for example, in saline environments [15,16]. The versatile roles Metarhizium spp. may have emphasize the beneficial potential and value of its presence, particularly in agricultural habitats [17]. Arable land, grassland, and forests represent major anthropogenic land-use types (LUTs) and cover about 40% of the global land area [18]. A number of studies have re- ported possible effects of LUTs on the presence and abundance of Metarhizium spp. Factores responsible includ crop types, farming practices management intensity, tillage, and the application of pesticides or fertilizers [19–22]. For instance, it has been consistently reported that Metarhizium spp. occur at higher abundances in unforested, sun-exposed LUTs or open land areas such as arable- and grasslands compared to forests, where their abundances are lower or sometimes not even detected [17,19,23–26]. The abundance of Metarhizium spp. has been more prominent in unforested semi-natural habitats such as permanent grassland and field margins compared to arable land in Switzerland [27,28]. Furthermore, there is evidence from field studies in Norway indicating a higher abundance of entomopathogenic fungi such as M. anisopliae, Beauveria bassiana (Bals.-Criv.) Vuill., or Tolypocladium cylindrospo- rum W. Gams in soils of organically managed fields compared to conventionally managed fields [29]. Several studies have suggested that the application of organic fertilizers may be the main driving factor for the increased abundance of Metarhizium spp. [22,30]. While the overall abundance of Metarhizium spp. in different LUTs has been explored in various studies, information on the species composition of Metarhizium communities and within-species diversity in different LUTs is limited. In addition, many of these studies have been performed prior to the establishment of the current Metarhizium species concept, which has precluded a comparison of data from these different periods. Nevertheless, the results reported by Cabrera-Mora et al. [31] have indicated that cropping systems can affect species abundances, as M. brunneum was isolated predominantly from soil where beans were grown and M. robertsii from soils where maize was grown. Similarly, Wyre- bek et al. [32] reported that M. robertsii, M. brunneum, and M. guizhouense were closely associated with soils from grassland, shrubs, or trees, respectively. In a study performed in Ontario, Canada, two genetically distinct groups were found to be associated with agricultural unforested field habitats and forested soils, respectively [23,33,34]. Isolates of the two genetic groups were subsequently described as M. brunneum and M. robertsii [35]. Studies assessing within-species diversity have mostly focused on individual fields or agricultural systems only, e.g., an agricultural field in Denmark [36] and a strawberry field in Brazil [37]. A comparison of the within-species diversity of M. brunneum and M. robertii between agricultural and adjacent natural habitats has been performed in different ecologi- cal regions in northwestern North America [38]. While considerable genotype diversity has been reported for both species in both habitats, no significant differences have been detected between them. Information on the number of species, their population structures, and distributions in different LUTs remains limited. This information, however, is impor- tant to understand the interaction of ecological conditions and Metarhizium population structures. In turn, it may allow the improvement of existing habitats or the development of new habitats and/or LUT-adapted biological control approaches, including conservation biological control [17]. Conditions in a particular soil or at a particular site are affected, on the one hand, by anthropogenic activities and, on the other hand, by different environmental factors, i.e., abiotic (physical and chemical), biotic (derived from organisms), as well as climatic (humidity, temperature, and solar radiation) factors. All these factors may have direct effects on the occurrence, abundance, propagation, persistence, and even virulence of Metarhizium populations [24,39–42]. Soil physical, chemical, and biological factors, such Microorganisms 2021, 9, 1380 3 of 19

as soil texture, pH, and organic matter content, belong to the most intensively studied factors in this context. However, various reports have not been able to provide consistent information, and, in many cases, they have been contradictory. For example, organic matter, which represents an important source of organic carbon in soil, has been reported to negatively correlate with the abundance of Metarhizium in agricultural land in Pennsylvania, USA [20], whereas other studies have reported positive correlations to agricultural land, grassland, and forests in Spain [24]. Similarly, the C:N ratio, a soil factor directly related to the quality of organic matter, has been reported to positively correlate with the abundance of entomopathogenic fungi in soil [43]. In that study, the C:N ratio was proposed as a predictor of Metarhizium spp. abundance in vineyard soils; however, whether this concept may generally be applicable to vineyards or even other habitats or LUTs such as grassland and arable land or even forests is currently not clear. The effect of soil pH on microbial populations is well documented; however, data on the specific effects of pH on Metarhizium abundance are currently not conclusive. While some studies suggest that Metarhizium prefer more alkaline environments, others have reported that Metarhizium are more adapted to acidic conditions or are not even responsive to pH [24,44]. In the present study, we determine Metarhizium spp. abundance and population structure at 30 different monitoring sites of the Swiss soil-monitoring network (NABO) that equally represent arable land, grassland, and forest. The overall goal was to obtain profound information on habitat selectivity and the factors that may drive population structures and the abundance of this important entomopathogenic fungal genus. Our specific goals were to: (1) determine Metarhizium spp. abundance as well as species diversity and population structures at the 30 sites across Switzerland, (2) investigate the differences among and within LUTs in detected Metarhizium populations, and (3) assess the extent to which soil physical, chemical, and biological factors may explain the differences observed.

2. Materials and Methods 2.1. Sampling Sites, Sampling Procedure The Swiss soil-monitoring network (NABO, www.nabo.ch), established in 1984, mon- itors 103 sites for physical, chemical, and biological parameters (biomass, soil respira- tion) [45]. The sites represent three different LUTs, i.e., arable land, permanent grassland, and forest, which dominate and characterize the landscape in the midlands and pre-alpine areas in Switzerland. In 2012, the NABO was complemented with the NABObio initiative to monitor soil microbial biodiversity at 30 NABO sites (10 sites of each of the three LUTs) (Figure S1) [46]. Sampling was carried out in spring after snowmelt and before the first fertilization, starting in Ticino (south of Switzerland) at the beginning of February 2016 and finishing in Grindelwald (alpine area) at the end of June 2016. This sampling scheme was implemented to harmonize time points in relation to the vegetation status of individual sites, i.e., the beginning of vegetation growth. Three composite samples were collected at each of the 30 sites. Composite samples consisted of 25 bulked soil cores (2.5 cm diameter × 20 cm depth) that were taken evenly distributed in a georeferenced 10 × 10 m plot, according to the standardized sampling protocol of NABO sampling [47]. In total, 90 composite samples were collected (Table S1) and transported to the laboratory immediately after sampling and kept at 4 ◦C until use. Thirteen soil factors (physical, chemical, and biological), for example, C:N ratio (organic carbon/total nitrogen), bulk density, and basal respiration, among others, were measured from each sample, as described previously by Gubler et al. [45]. In addition, mean annual temperature (MAT) and mean annual precipitation (MAP) were determined based on yearly data from 1981 to 2015 from MeteoSwiss (http://www.meteoswiss.admin. ch). In total, 15 environmental factors were evaluated in this study (Table S2). Microorganisms 2021, 9, 1380 4 of 19

2.2. Fungal Isolation and DNA Extraction Metarhizium abundance was determined, and single colonies were isolated from each of the three mixed soil samples of the 30 sites. The samples were sieved with a 2 mm mesh, and soil water content was determined. Then, 5 g of each homogenized sample was suspended in 25 mL of sterile distilled water + 0.1% Tween 80 in a 100 mL Erlenmeyer flask. Flasks were stirred in a rotatory shaker at 120 rpm at room temperature for 3 h. After 20 s of sedimentation, 100 µL aliquots of the suspension were spread on 90 mm Petri dishes containing a semi-selective medium (SM) with dodine (Discovery, Leu Gygax AG, Switzerland) as the selective compound [48]. Two replicates were plated per soil sample. Cultures were grown for 15 d at 24 ± 1 ◦C in the dark and inspected every 2 days. The fungal isolates were classified to the genus level using taxonomic keys [49]. Metarhizium colony forming units (CFU) per g (dry weight) of soil were determined for each plate and means calculated per site. Three sporulating colonies were randomly taken per plate. Conidia of each colony were plated with a dilution smear in a new SM plate, and a monosporic colony was isolated. In total, a maximum of 18 isolates was collected per site. All the fungal isolates obtained were deposited in the culture collection of the Molecular Ecology Group, Agroscope, Zürich, Switzerland. Monosporic cultures were plated onto sterile filter paper placed on potato dextrose agar (PDA) plates to produce mycelium for DNA extraction. After 4 days of incubation at 24 ± 1 ◦C, mycelium was scraped off the filter paper with a sterile scalpel, placed in an Eppendorf tube, and lyophilized for six hours at −4 ◦C using a CentriVap benchtop centrifugal vacuum concentrator (LabConco, Kansas City, MO, USA). Genomic DNA of each isolate was extracted using the NucleopSpin Plant II kit (Machery & Nagel, Düren, Germany). DNA concentration was determined using a Cary Elipse fluorescence spec- trophotometer (Varian, Palo Alto, CA, USA) with a Picogreen® fluorescent nucleic acid stain (Invitrogen, Carlsbad, CA, USA). The DNA extracts were standardized to 5 ng µL−1.

2.3. Sequence Analysis Metarhizium isolates were taxonomically identified to the species level by sequencing the 50 end of the nuclear translation elongation factor-1α (50-TEF-1α) and subsequent alignment with reference sequences. The 50-TEF region was amplified using primers EF1T 50-ATGGGTAAGGARGACAAGAC-30 [50] and EFjmetaR 50-TGCTCACGRGTCTGGCCAT CCTT-30 [51]. PCRs were performed in 20 µl reaction volumes consisting of 15 ng genomic DNA and 1x Phusion HF buffer containing 7.5 mM MgCl2, 0.2 mM dNTPs, 0.2 µM of each primer, 3% dimethyl sulfoxide (DMSO), and 0.2 U of Phusion Hot Start II High-Fidelity DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA). PCR conditions were: 30 s denaturation at 98 ◦C, followed by 38 cycles of 5 s denaturation at 98 ◦C, 20 s annealing at 58 ◦C, and a 1 min extension at 72 ◦C. Reactions were completed with a 10 min elongation at 72 ◦C. PCR products were cleaned up using the NucleoSpin® Gel and PCR Clean-up kit (Machery & Nagel, Düren, Germany), following the manufacturer’s protocol. Sequencing reactions were performed with primers EF1T and EFjmetaR using the BigDye™ Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA, USA) and sequences deter- mined using an ABI PRISM 3130xl genetic analyzer, as described above. Sequences were

assembled and corrected using the software DNABaser V.4 (Heracle Biosoft, Pites, ti, Roma- nia). The obtained sequences were aligned with reference sequences of Metarhizium spp. type isolates for species allocation, as described in Mayerhofer et al. [51]. The sequences of one isolate per multilocus genotype (MLG) are deposited at GeneBank (accession num- bers MZ297396 to MZ297417, AppendixA). Reference sequences were downloaded from GenBank, and alignments were performed using the Clustal-W subsequence realignment tool implemented in MEGA X [52]. The maximum likelihood method based on the Kimura 2-parameter model implemented in MEGA X was used to calculate (1000 iterations) and construct a phylogenetic tree [53]. Microorganisms 2021, 9, 1380 5 of 19

2.4. Multilocus Microsatellite Genotyping Fourteen microsatellite markers were used in five primer pair sets, as described by Mayerhofer et al. [54], to assess the genotype of the collected Metarhizium isolates. Multiplex PCRs were performed in 20 µL reaction volumes containing 5 ng of genomic DNA, 1× GoTaq colorless Flexi Buffer, 0.2 mM of each primer (forward primer labeled with FAM, HEX, or NED), 0.2 mM of dNTPs, 3 or 4 mM of MgCl2 (depending on the multiplex set, [54]) and 0.25 U of GoTaq G2 Flexi DNA Polymerase (Promega, Madison, WI, USA). Cycling conditions included an initial denaturing step of 2 min at 94 ◦C, followed by 12 touchdown cycles of 30 s denaturation at 94 ◦C, 30 s annealing at Ta + 12 ◦C, (with a 1 ◦C decrease per cycle), and 40 s of extension at 72 ◦C. This was followed by another 22 or 30 cycles [54] of 30 s denaturation at 94 ◦C, 30 s at Ta [54], and a 40 s extension at 72 ◦C. Reactions were completed with a 15 min elongation at 72 ◦C. PCR product sizes (microsatellite allele size) were determined on an ABI 3500 Series genetic analyzer (24 capillaries) using POP-7 polymer (Applied Biosystems, Foster City, CA, USA). GenScan ROX400 (Applied Biosystems) was used as an internal size standard. Data were analyzed using GenMarker V2.4.0 (SoftGenetics, State College, PA, USA) and allele sizes corrected according to fragment sizes of reference strains M. brunneum ARSEF7524 and M. robertsii ARSEF7532.

2.5. Data Analysis Analysis of variance (ANOVA) was used to assess the effect of LUT, site, and soil environmental factors on the abundance of Metarhizium (CFU g−1 of soil dry weight), followed by the Tukey-HSD test to assess pairwise differences. Sites with zero Metarhizium spp. abundance were not included in the calculations of differences among LUTs to provide a clear representation of mean fungal abundance where Metarhizium was detected. Microsatellite marker data analyses were performed with the R package Poppr [55]. Data were clone-corrected, and the number of MLGs calculated. The Shannon-Wiener index (H), [56] and the evenness index (E.5) [57–59] were calculated for each Metarhizium species and LUT based on the occurrence of each MLG. Differences in the population structure of Metarhizium MLGs among LUTs and species were assessed with overall and pairwise PERMANOVA based on Bray–Curtis (BC) dis- similarity matrices [60] using the function “adonis” within the vegan R package, “pair- wise.perm.manova” within the RVAideMemoire R package [61,62], and the Benjamini– Hochberg p-value correction in R [63]. Effects of soil factors on Metarhizium populations were assessed with overall PERMANOVA tests for each factor individually using the “adonis” function. Principal coordinate analyses (PCoA) were performed based on BC dissimilarity matrices using the “cmdscale” function included in the R core package [63,64]. The “envfit” function from the vegan R package was used to plot the correlations between Metarhizium species population structures and the significant soil factors determined in a PCoA ordination.

3. Results 3.1. Abundance of Metarhizium in Soil Using a semi-selective medium, the genus Metarhizium was detected at eight of ten arable land sites, at all ten permanent grassland sites, and at four of ten forest sites (Figure1 and Table S3). LUT significantly (F2,19 = 4.30; p-value = 0.0287) affected the abundance of Metarhizium, determined as CFU g−1 soil dry weight. Pairwise tests (Tukey-HSD test, p-value < 0.05) revealed significant differences between the site means of grassland (3501.2 ± 2254.2 CFU g−1 soil dry weight) and arable land (1199.0 ± 1278.5 CFU g−1 soil dry weight) but not between the means of grassland and forest (1620.8 ± 938.2 CFU g−1 soil dry weight) or arable land and forest. Significant differences (F29,58 = 11.63; p-value < 0.0001) in Metarhizium abundances were also detected among sites (Figure1 and Table S3), where abundance ranged from 0 to 6362.5 ± 1467.2 CFU g−1 soil dry weight (mean ± SD). Microorganisms 2021, 9, x FOR PEER REVIEW 6 of 19

Metarhizium abundances at the 30 sites were grouped into three distinct groups, i.e., sites with low abundance (mean abundance < 150 CFU g−1 soil dry weight), medium abun- dance (mean abundance ranging between 150 and 4000 CFU g−1 soil dry weight), and high abundance (sites with a mean abundance > 4000 CFU g−1 soil dry weight) (Figure 1 and Table S3). Medium Metarhizium abundance was detected at sites of all three LUTs (14 sites), whereas high abundance was only detected in grassland (5 sites). The low abun- dance group included sites where Metarhizium was detected with a low abundance (3 sites in arable land) as well as sites where Metarhizium was not detected (2 sites in arable land, and 6 sites in forest). Arable land was the only LUT that included sites with low abun- Microorganisms 2021, 9,dance 1380 as well as sites where the fungus was not detected (5 sites). However, due to miss- 6 of 19 ing statistical power, no discrimination of these sites was possible, and they were all com- bined in the group of low abundance.

Figure 1. MetarhiziumFigure 1.spp. Metarhizium abundance spp. represented abundance as represented dot plots of as colony-forming dot plots of colony-forming units (CFU g− units1 of soil (CFU dry g− weight)1 of for −1 different sites (30soil NABObiodry weight) sites). for different Dots represent sites (30 values NABObio of CFU sites). g−1 Dotsof soil represent dry weight values in eachof CFU sample g of per soil site, dryand the −1 intersection representsweight in the each mean sample of CFU per site, g−1 andof soil the dry intersection weight per represents site. * Dashed the mean line of (4000 CFU CFUg of gsoil−1 ofdry soil weight dryweight) per site. * Dashed line (4000 CFU g−1 of soil dry weight) separates the high abundance group from separates the high abundance group from the medium abundance group; ** Dashed line (150 CFU g−1 of soil dry weight) the medium abundance group; ** Dashed line (150 CFU g−1 of soil dry weight) separates the medium separates the medium abundance group from the low abundance group) (Table S3). Site numbers where Metarhizium was abundance group from the low abundance group) (Table S3). Site numbers where Metarhizium was not detected arenot underlined. detected are underlined. Metarhizium abundances at the 30 sites were grouped into three distinct groups, i.e., 3.2. Effect of Environmental Factors on Metarhizium Abundance Groups sites with low abundance (mean abundance < 150 CFU g−1 soil dry weight), medium In total,abundance 15 environmental (mean abundance factors were ranging measured between at the 150 30 and sites 4000 and CFU assessed g−1 soil for dry dif- weight), and ferences amonghigh abundancethe sites categorized (sites with aaccording mean abundance to the three > 4000 Metarhizium CFU g−1 soil abundance dry weight) (Figure groups (Table1 and 1 and Table Tables S3). S4–S7). Medium TwoMetarhizium factors, theabundance C:N ratio and was mean detected annual at sitesprecipita- of all three LUTs tion (MAP),(14 differed sites), significantly whereas high among abundance the Metarhizium was only abundance detected ingroups grassland independent (5 sites). The low of the three abundanceLUTs between group high included and low sites as wherewell asMetarhizium medium andwas low detected abundance with agroups low abundance (3 but not betweensites high in arable and medium land) as wellabundance as sites gr whereoups Metarhizium(Table 1 and wasTable not S4). detected Within (2each sites in arable LUT, differentland, factors and 6were sites significant in forest). be Arabletween land the waslow, themedium, only LUT and thathigh included abundance sites with low groups. In arableabundance land, the as wellsix soil as sitesfactors—al wheretitude, the fungus clay, wassand, not soil detected skeleton, (5 total sites). carbon, However, due to and DNA (proxymissing for statisticalbiomass)—differed power, no signif discriminationicantly between of these the sites lowwas and possible,medium andabun- they were all dance groupscombined (Table 1 inand the Table group S5). of In low grassl abundance.and, only silt significantly differed between high and medium abundance, whereas in forest, altitude, clay, sand, and C:N ratio signif- icantly differed3.2. Effectbetween of Environmental medium and Factorslow Metarhizium on Metarhizium abundance Abundance (Tables Groups 1, S6, and S7). In total, 15 environmental factors were measured at the 30 sites and assessed for differences among the sites categorized according to the three Metarhizium abundance groups (Table1 and Tables S4–S7). Two factors, the C:N ratio and mean annual precipitation (MAP), differed significantly among the Metarhizium abundance groups independent of the three LUTs between high and low as well as medium and low abundance groups but not between high and medium abundance groups (Table1 and Table S4). Within each LUT, different factors were significant between the low, medium, and high abundance groups. In arable land, the six soil factors—altitude, clay, sand, soil skeleton, total carbon, and DNA (proxy for biomass)—differed significantly between the low and medium abundance groups (Table1 and Table S5). In grassland, only silt significantly differed between high and medium abundance, whereas in forest, altitude, clay, sand, and C:N ratio significantly differed between medium and low Metarhizium abundance (Table1, Tables S6 and S7). Microorganisms 2021, 9, 1380 7 of 19

Table 1. Summary of soil and environmental factors that differ significantly among Metarhizium spp. abundance groups (high, medium, and low) and between and within land-use types (LUTs).

High Medium Low ANOVA LUT Factor (1) Mean SD Minimum Maximum Mean SD Minimum Maximum Mean SD Minimum Maximum F-Value p-Value Pattern (2) C:N ratio 9.6 0.9 8.9 10.0 11.1 3.1 8.0 17.7 14.2 5.5 8.6 27.0 8.72 0.0004 H = M 6= L All LUT MAP [mm] 1496.2 321.0 1090.0 1910.0 1342.4 363.2 962.0 2140.0 1116.7 368.5 528.0 1838.0 6.77 0.0019 H = M 6= L Altitude [masl] - (3) - - - 549.4 167.9 336.0 830.0 449.6 55.6 379.0 545.0 4.78 0.0374 M 6= L Clay [% sdw] - - - - 17.6 4.6 11.5 23.8 32.5 19.1 5.8 59.0 8.69 0.0064 M 6= L Arable Sand [% sdw] - - - - 42.2 9.8 31.0 54.0 22.4 11.4 11.0 36.1 26.15 <0.0001 M 6= L land Soil skeleton [% - - - - 4.0 0.8 3.1 4.9 0.7 0.8 0.0 2.0 135.85 <0.0001 M 6= L sv] Total Carbon [% - - - - 2.0 0.7 1.1 3.2 2.8 0.9 1.8 4.3 8.55 0.0068 M 6= L sdw] DNA [µg/mg - - - - 25.5 9.1 14.0 45.0 17.9 3.8 13.0 26.0 8.85 0.006 M 6= L sdw] Grassland Silt [% sdw] 42.3 8.4 34.3 55.0 35.6 7.9 27.0 49.9 - - - 5.10 0.0319 H 6= M Altitude [masl] - - - - 745.3 269.9 525.00 1180.0 1115.8 408.7 505.0 1655.0 7.60 0.0101 M 6= L Clay [% sdw] - - - - 31.1 9.1 18.8 42.0 17.8 8.3 7.0 30.5 17.30 0.0003 M 6= L Forest Sand [% sdw] - - - - 31.1 8.7 18.0 39.0 47.2 21.0 17.5 71.0 6.35 0.0177 M 6= L C:N ratio - - - - 15.5 2.2 12.0 17.7 18.2 4.3 13.7 27.0 4.25 0.0488 M 6= L (1) Environmental factors that revealed significant differences are shown. For the complete analysis of the 15 environmental factors evaluated, see Tables S4–S7. MAP: mean annual precipitation (1981 to 2015); masl: meters above sea level; sdw: soil dry weight; sv: soil volume. (2) Significant differences in pairwise tests between the proportion of colony-forming units (CFU) (p < 0.05, Tukey-HSD test). (3) Metarhizium was not detected at this abundance level. Microorganisms 2021, 9, 1380 8 of 19

3.3. Metarhizium Species Occurrence and Genetic Diversity 3.3. Metarhizium Species Occurrence and Genetic Diversity In total, 349 isolates were collected from the 22 sites at which Metarhizium was detected and, basedIn total, on 349 sequence isolates analyses were collected of the 5 0fromend ofthe nuclear 22 sites translation at which elongationMetarhizium factor-1 was de-α (5tected0-TEF-1 and,α), based assigned on sequence to three Metarhizium analyses of species:the 5′ endM. of brunneum nuclear translation(259 isolates), elongationM. robertsii fac- (80tor-1 isolates),α (5′-TEF-1 andα),M. assigned guizhouense to three(10 isolates)Metarhizium (Table species:2 and M. Table brunneum S1). M. (259 brunneum isolates),was M. presentrobertsiiin (80 all isolates), three LUTs and (7 M. arable guizhouense land sites, (10 10isolates) grassland (Tables sites, 2 and 4 forest S1). sites),M. brunneum whereas wasM. robertsiipresent andin allM. three guizhouense LUTs (7 werearable present land sites, in arable 10 grassland land (M. sites, robertsii 4 forest6 sites, sites),M. guizhouensewhereas M. 1robertsii site) and and grassland M. guizhouense only (M. were robertsii present4 sites, in arableM. guizhouense land (M. robertsii2 sites; 6 Table sites,2 andM. guizhouenseFigure2 ). The1 site) proportion and grassland of M. only brunneum (M. robertsiiand M. 4 robertsiisites, M.isolates guizhouense was 2 similar sites; Table in the 2 lowand (61%Figure and 2). 39%)The proportion and medium of M. (56% brunneum and 40%) and abundance M. robertsii groups isolates in was arable similar land in and thethe low high (61% (57% and and39%) 36%) and abundancemedium (56% group and in 40%) grassland abundance (Figure groups2). In in contrast, arable inland the and medium the high abundance (57% and group36%) abundance in grassland, group the portion in grassland of M. brunneum(Figure 2)isolates. In contrast, was higher, in the i.e., mediumM. brunneum abundance94% andgroupM. in robertsii grassland,5% (Figurethe portion2). M.of M. guizhouense brunneum wasisolates detected was higher, in the i.e., medium M. brunneum abundance 94% groupand M. in robertsii arable land 5% (4%)(Figure and 2). grassland M. guizhouense (1%) and was the detected high abundance in the medium group in abundance grassland (7%)group (Figure in arable2). land (4%) and grassland (1%) and the high abundance group in grassland (7%) (Figure 2). Table 2. Numbers of isolates, multilocus genotypes, and genotype diversity among the three Metarhizium species—M. brunneumTable 2. ,NumbersM. robertsii of, andisolates,M. guizhouense multilocus—for genotypes, three different and genotype LUTs:arable diversity land, among grassland, the three and forest.Metarhizium species—M. brunneum, M. robertsii, and M. guizhouense—for three different LUTs: arable land, grassland, and forest. Metarhizium spp. M. brunneum M. robertsii M. guizhouense Metarhizium spp. M. brunneum M. robertsii M. guizhouense Land-Use Type N MLG N MLG H E.5 N MLG H E.5 N MLG H E.5 Land-Use Type N MLG N MLG H E.5 N MLG H E.5 N MLG H E.5 ArableArable land land 115 115 10 10 66 4 66 0.99 4 0.750.99 0.7546 46 5 5 1.09 1.09 0.600.60 3 3 1 - - - - GrasslandGrassland 171 171 12 12 130 1307 1.33 7 0.751.33 0.7534 34 3 3 0.70 0.70 0.740.74 7 7 2 - - - - ForestForest 6363 44 63 463 0.964 0.800.96 0.80 ------Total 349 22 259 13 1.84 0.72 80 6 1.17 0.74 10 3 - - Total 349 22 259 13 1.84 0.72 80 6 1.17 0.74 10 3 - - N: totalN: number total number of isolates; of isolates; MLG: MLG: number number of of unique unique multilocusmultilocus genotypes; genotypes; H: Shannon–WienerH: Shannon–Wi index;ener index; E.5: evenness. E.5: evenness.

FigureFigure 2.2.Number Number of of isolates isolates recovered recovered for for the the three three species species identified identified (M. ( brunneumM. brunneum, M., robertsii M. robertsii, and, M.and guizhouense M. guizhouense) in each) in each of the of three the three land-use land-use types—arable types—arab landle (AL),land (AL), grassland grassland (G), and (G), forest and forest (F)— separated(F)—separated according according to the threeto the abundance three abundance groups, gr lowoups, (L), mediumlow (L), medium (M), and high(M), (H)and (see high Figure (H) (see1). Figure 1). Genotyping of the isolates based on microsatellite analyses revealed 22 different MLGs amongGenotyping the 349 isolates: of the 13isolates MLGs based for M. on brunneum microsatellite, 6 MLGs analyses for M. robertsiirevealed, and 22 3different MLGs forMLGsM. guizhouenseamong the ,349 revealing isolates: that 13 theseMLGs microsatellite-based for M. brunneum, 6 MLGsMLGs werefor M. species-specific robertsii, and 3 (FigureMLGs 3for and M. Table guizhouense2). Overall,, revealing the number that these of detected microsatellite-based MLGs was similar MLGs in were arable species- land and grassland, while it was 2.5 to 3 times lower in forest. For M. brunneum, the number of

Microorganisms 2021, 9, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/microorganisms Microorganisms 2021, 9, x FOR PEER REVIEW 2 of 19 Microorganisms 2021, 9, 1380 9 of 19

specific (Figure 3 and Table 2). Overall, the number of detected MLGs was similar in ara- MLGsble land was and higher grassland, in grassland while comparedit was 2.5 to to 3 arable times landlower and in forest. forest, For whereas M. brunneum for M robertsii, the , thenumber number of ofMLGs MLGs was was higher higher in grassland in arable co land.mpared Shannon–Wiener to arable land and genetic forest, diversity whereas (H) forforM. M brunneum robertsii, thewas number highest of in MLGs permanent was higher grassland, in arable followed land. Shannon–Wiener by arable land and genetic forest, whilediversity for M. (H) robertsii for M., diversitybrunneum was was higher highest in in arable permanent land and grassland, lower infollowed grassland by (Tablearable2 ). Inland contrast, and forest, the evenness while for (E.5) M. ofrobertsiiM. brunneum, diversitywas was higher higher in in forest arable compared land and to lower grassland in andgrassland arable land,(Table and 2). In the contrast, evenness the of evennessM. robertsii (E.5)was of M. higher brunneum in grassland was higher compared in forest to compared to grassland and arable land, and the evenness of M. robertsii was higher in arable land (Table2). Due to the low number of M. guizhouense isolates identified, diversity grassland compared to arable land (Table 2). Due to the low number of M. guizhouense indices for this species were not calculated. isolates identified, diversity indices for this species were not calculated.

FigureFigure 3. Maximum 3. Maximum likelihood likelihood phylogenetic phylogenetic tree tree based based on on the the alignment alignment of of 50 5-TEF-1′-TEF-1ααsequences sequences representing representingeach each ofof thethe 22 22 different multilocus genotypes (MLG). Bootstrap values > 60%, calculated with 1000 replicates, are shown. The bar scale different multilocus genotypes (MLG). Bootstrap values > 60%, calculated with 1000 replicates, are shown. The bar scale indicates 0.01 changes per nucleotide. Each circle represents a unique MLG, and circle size indicates the number of isolates recovered. Colors within the circles represent the fraction of clones obtained per land-use type, and vertex size represents the fraction of clones isolated at sites where the MLG was detected. Microorganisms 2021, 9, 1380 10 of 19

In total, 18 of the 22 MLGs were specific to one of the three LUTs. Six MLGs were specific to arable land (two M. brunneum, three M. robertsii, one M. guizhouense), eight to permanent grassland (five M. brunneum, one M. robertsii, two M. guizhouense), and four to forest, which clustered separately within M. brunneum in the 50-TEF-1α-based phylogenetic tree (Figure3). None of the MLGs were detected in all three LUTs, while the four most abundant MLGs, i.e., two M. brunneum (15 and 19) and two M. robertsii MLGs (2 and 13), were present in arable land and grassland and represented 63.61% of all Metarhizium isolates collected (Figure3). The four forest-specific MLGs represented 18.05% of all Metarhizium spp. isolates investigated. The abundance of individual MLGs did not correlate with affiliation to particular Metarhizium abundance groups, e.g., M. brunneum MLG15 and M. robertsii MLG 13 were present in the low, medium, and high abundance groups in different sites (data not shown). Eleven MLGs (five M. brunneum, three M. robertsii, and three M. guizhouense) were isolated from single sites only (Figure3). In contrast, MLG 15 comprised isolates from 15 different sites. Two MLGs (MLG 4 and 5) of the four forest-specific MLGs included isolates from all the four sites and two included isolates from two sites (MLG 6) and one (MLG 10) site where M. brunneum was detected (Figure3).

3.4. Land-Use Type Effect on Metarhizium Populations Overall, PERMANOVAs indicated that the three LUTs significantly explained the vari- ation in M. brunneum population structures (R2 = 0.37; pseudo-F = 5.36; p-value = 0.0001) (Table3 ). PCoA explained 47.00% of the observed variation among M. brunneum popula- tions, i.e., 24.40% of the variation in the first axis, separating forest populations from arable land and grassland populations and 22.60% of the variation in the second axis, differentiat- ing between arable land and grassland populations (Figure4A). Pairwise PERMANOVA tests among the three LUTs revealed significant differences between M. brunneum pop- ulations from forest and arable land (p-value = 0.003), between forest and grassland (p-value = 0.003), but not between grassland and arable land (p-value = 0.05). Overall, PERMANOVA indicated that the LUT (arable land, grassland) did not significantly explain the variation among M. robertsii population structures (p-value = 0.466) (Table3 ). PCoA explained 70.40% of the observed variation among M. robertsii arable land and grassland populations, i.e., 52.00% of the variation in the first axis and 28.40% of the variation in the second axis (Figure4B). Variations among M. guizhouense populations were not evaluated further due to the low number of isolates (10) and genotypes (3 MLGs) detected.

Table 3. Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors, which significantly affect M. brunneum and M. robertsii population structures among and within each land-use type. Analyses were based on Bray–Curtis distances.

M. brunneum M. robertsii Grassland– Among Arable Between 2 Arable Arable Grassland Forest Grassland 3 LUTs Land LUTs Land Land p-Value p-Value p-Value p-Value p-Value p-Value p-Value p-Value Variable (R2) (R2) (R2) (R2) (R2) (R2) (R2) (R2) Land-use type 0.0001 (37%) 0.0513 - - - 0.4665 - - C:N ratio 0.0001 (25%) 0.1149 0.9821 0.0074 (26%) 0.2083 0.5366 0.4599 0.8333 Basal respiration 0.0071 (13%) 0.0928 0.3016 0.7186 0.1667 0.1327 0.4595 0.6668 Organic Carbon 0.0039 (12%) 0.0861 0.5563 0.6421 0.2083 0.2562 0.6476 0.8333 Total Carbon 0.0055 (11%) 0.1657 0.5891 0.6488 0.2083 0.1403 0.3230 0.8333 Bulk density 0.0055 (13%) 0.2375 0.5190 0.0893 0.2500 0.1333 0.6103 0.1667 Soil skeleton 0.5821 0.4503 0.8032 0.5633 0.8333 0.0098 (33%) 0.3008 0.3333 R2 values are shown in parentheses when significant (in bold). For the complete analysis, see Tables S8–S15. Microorganisms 2021, 9, x FOR PEER REVIEW 4 of 19

Bulk density 0.0055 (13%) 0.2375 0.5190 0.0893 0.2500 0.1333 0.6103 0.1667 Microorganisms 2021, 9, 1380 11 of 19 Soil skeleton 0.5821 0.4503 0.8032 0.5633 0.8333 0.0098 (33%) 0.3008 0.3333 R2 values are shown in parentheses when significant (in bold). For the complete analysis, see Tables S8–S15.

FigureFigure 4.4. PrincipalPrincipal coordinatecoordinate analysisanalysis (PCoA)(PCoA) basedbased onon aa Bray–CurtisBray–Curtis dissimilaritydissimilarity matrixmatrix showingshowing differencesdifferences betweenbetween ((AA)) M.M. brunneumbrunneum andand ((BB)) M.M. robertsiirobertsii populationspopulations atat thethe sitessites ofof threethree land-useland-use types:types: arablearable land,land, grassland,grassland, and forest. CirclesCircles (arable(arable land),land), squaressquares (grassland),(grassland), andand trianglestriangles (forest)(forest) representrepresent M.M. brunneumbrunneum oror M.M. robertsiirobertsii populationspopulations atat thethe different sites (site numbers indicated). Convex hulls enclose samples of the same land-use type. The percentage of variation explained by the plotted principal coordinates is indicated on the axes. Vectors represent the best fitted and explained environmental factors, and vector lengths represent the strength of correlation. C:N ratio: ratio of organic carbon and total nitrogen; Ctot: total carbon; Corg: organic carbon; BR: basal respiration; BD: bulk density; SS: soil skeleton. Microorganisms 2021, 9, 1380 12 of 19

3.5. Effect of Environmental Factors on Metarhizium Populations PERMANOVA revealed that 5 of the 15 environmental factors significantly explained the variation of M. brunneum population structures among the three LUTs (Table3 and Figure4A). Most variation was explained by the factor C:N ratio (25%), whereas basal respiration, total carbon, organic carbon, and bulk density explained the variation in the same range of 11% to 13%. Soil skeleton was the only factor that significantly explained the variation among M. robertsii populations from arable land and grassland (Table3 and Figure4B). Except for the factor C:N ratio, which significantly explained the variation of M. brunneum populations among grassland sites, no other environmental factor significantly explained the variation of M. brunneum or M. robertsii populations within arable land, grassland, or forest.

4. Discussion Identification of the factors that drive population structure and development of insect pathogenic fungi such as Metarhizium spp. is a fundamental requirement to exploit their potential in biological control. The results presented in this study emphasize the effect and influence of particular LUTs on Metarhizium abundance and population structures and indicate the complexity of the factors involved. Populations at the 30 sites investigated were composed of the three species M. brunneum, M. robertsii, and M. guizhouense, of which M. brunneum (with 74% of isolates recovered) was the most abundant and genotypically diverse species and M. guizhouense (with 3%) the least abundant of all isolates collected. Distinct differences in Metarhizium population structures were detected between forest and the two unforested LUTs, arable land and grassland, revealed by the finding that M. robertsii and M. guizhouense were not detected in forest and M. brunneum was represented by four forest-specific genotypes. Although population structures did not differ between the two unforested LUTs, Metarhizium spp. abundance was significantly higher in grassland compared to arable land.

4.1. M. brunneum Populations in Forested vs. Unforested LUTs The forest sites of the NABO network provided characteristic soil environmental con- ditions, with reduced exposure to sunlight, low soil bulk density, and high basal respiration, as well as high carbon content and C:N ratio (Table S16). These physical, chemical, and biological factors were correlated with M. brunneum population variation among the 10 forest sites and 20 arable land or grassland sites, indicating their importance to the structure of Metarhizium populations. The difference between M. brunneum populations in forested and unforested sites was explained particularly by the soil factor C:N ratio, which was also the only factor that significantly explained M. brunneum population variation within one LUT, i.e., among grassland sites. The C:N ratio also differed significantly between low and medium Metarhizium abundance groups among all LUTs and among forest sites (Table 1). Our results suggest that Metarhizium abundance and species and genotype diversity tend to be higher when the C:N ratio is medium–low (Tables1–3). In contrast to our results, Uzman et al. [43] reported enhanced Metarhizium spp. abundance with increasing C:N ratios in vineyard soils, and Clifton et al. [22] showed a negative correlation between Metarhizium spp. abundance and N concentration, a factor indirectly related to C:N ratios in soils of conventional and organically farmed fields. The latter authors hypothesized that this effect may be due to a stimulation of certain microorganisms that exploit elevated N concentrations and, subsequently, may outcompete Metarhizium spp. In vitro studies have demonstrated that the proportion of C and N in culture media considerably affects the production of biomass, conidia, and insecticidal molecules as well as the virulence of Metarhizium spp. [65–67]. These laboratory experiments were performed with a variety of C and N sources in artificial media, and, therefore, it remains unresolved as to what extent these findings may apply to soil. Microorganisms 2021, 9, 1380 13 of 19

4.2. Metarhizium Populations in Arable Land vs. Grassland Although M. brunneum and M. robertsii population structures in arable land and grassland were not significantly different, we detected significant differences between the two LUTs in overall abundance as well as the abundance group associations of the sites. Mean CFU g−1 values in the grassland sites were approximately twice as high as the values in arable land. In addition, in grassland, Metarhizium was present with high or medium abundance at all the sites, whereas in arable land, Metarhizium was present with medium and low abundances and not even detected at two sites. Interestingly, at the forest sites, the fungus was detected at only 4 out of 10 sites, all belonging to the medium abundance group. The mean abundance was similar to the one detected in arable land when considering only sites where the fungus was detected. MAP is a factor that characterizes the different sites in terms of precipitation. It has been shown to correlate with soil humidity [68], which is well recognized as a critical factor for fungal abundance and the colonization of soil (e.g., [69]). The fact that the MAP values in our study were significantly higher at sites with medium abundance compared to sites with low Metarhizium abundance indicates that the MAP in our system also affected soil humidity, which, in turn, induced Metarhizium growth. However, we did not find any within the LUT correlation of MAP with Metarhizium abundance. Other factors that were significantly correlated with Metarhizium abundance groups within different LUTs, e.g., altitude, clay, or sand, did not reveal significant differences among LUTs, which may indicate that these factors characterize geographical differences among the sites rather than LUTs. Direct comparisons of different studies are generally difficult to perform due to differ- ences in isolation techniques, sampling schemes, habitat definitions, geographic and/or climatic zones. Additional local factors such as climat, vegetation, and geographic his- tory may increase the complexity of interacting factors. However, in accordance with our results, Keller et al. [27] also reported a higher abundance of Metarhizium spp. (M. anisopliae complex) in meadows (semi-managed unforested land) compared to agricul- tural land in Switzerland. Furthermore, Schneider et al. [28] detected higher Metarhizium clade1 (M. anisopliae species complex) ITS copy numbers using real-time PCR in permanent grasslands and improved field margins compared to arable land. Such unforested land types do not receive pesticide treatments and are not subject to soil disturbances such as tillage. A number of studies have shown that soil receiving reduced or no soil manage- ment, including organically farmed soils, will allow a higher abundance of Metarhizium or entomopathogenic fungi in general [22,29,30,70,71]. Interestingly, it has been shown that, in contrast, tillage may also increase Metarhizium spp. abundance [20,21]. The authors of these studies have argued that tillage may disperse Metarhizium propagules, resulting in a more homogenous distribution of the fungus in the soil and, thus, increasing its overall abundance.

4.3. Association Metarhizium Species and Genotypes to LUTs We observed different affinities of M. brunneum, M. robertsii, and M. guizhouense to the three LUTs. While M. brunneum was detected in all LUTs (73.3% of the sites) and was the dominant species at 16 out of 22 sites as well as in the three abundance groups, M. robertsii and M. guizhouense were detected in arable land and grassland only (36.6% and 10.0% of the sites). However, whereas M. robetsii was detected within low and medium abundance groups in arable land and low and high abundance groups in grassland, M. guizhouense was detected only in the medium abundance group in arable land and medium and high abundance groups in grassland. The M. brunneum genotypes detected at the forest sites were specific to forests and clustered separately from the genotypes detected at the arable land or grassland sites. In addition, certain MLGs were particularly dominant and present at multiple sites across the country (MLG2, MLG13, MLG 15, MLG 19). Species affinity to particular sites, regions, or habitats and the dominance of particular genotypes have been reported previously on a field level [21,36] as well as on a regional level [38,72,73]. Plant species composition characteristically differs between habitats or LUTs such as agricultural Microorganisms 2021, 9, 1380 14 of 19

land, grassland, and forest. Previous studies have provided evidence that Metarhizium spp. can form stable associations with certain plant species [32,74], and that crop type may affect Metarhizium spp. diversity and abundance [21,31]. For instance, Cabrera- Mora et al. [31] predominantly isolated M. brunneum from bean-cultivated soil and M. robertsii from maize-cultivated soil, and Kepler et al. [21] detected higher Metarhizium abundance in soybean fields versus cornfields. Furthermore, it is well documented that populations vary substantially among LUTs [75]. Species associated with forests, for instance, bark (Scolytus spp.), may not be present in unforested habitats, and pest species like wireworm spp. or western corn rootworm are linked to certain crop plants. The population structure of arthropod pathogenic fungi such as Metarhizium, which includes species and genotypes with diverse host ranges (from narrow to broad), is likely affected and shaped by arthropod populations (host selection). Furthermore, it has been observed that small soil arthropods, such as collembolans and mites, may be involved in the dispersal of fungal propagules in the soil, which may also have an effect on fungal population development [76]. The data reported here and observations made by others strongly indicate the importance of plant and arthropod populations in particular habitats and LUTs as the driving force for shaping the population structure of entomopathogenic fungi such as Metarhizium spp.

5. Conclusions The 30 sites investigated in this study represent 3 well-defined and characterized habitat types in Switzerland. Being part of the Swiss soil-monitoring network (NABO), the sites have been cultivated, maintained, and monitored for the last three decades according to the same procedures and protocols, and therefore, represent well-established habitats. This long-term experimental system allowed us to unravel the significant impacts of the different LUTs on Metarhizium abundance, species diversity, and their population structures and indicated the complexity of environmental factors involved. With the detection of LUT-specific MLGs of M. brunneum and the absence of M. robertsii and M. guizhouense at forest sites, the study has emphasized the significant impact of forested versus unforested LUTs on Metarhizium population structures. These results will provide an important base to further explore other factors that may shape Metarhizium populations, which, for instance, may include the effects of plant populations as well as the diversity of arthropod populations present in the soil. Information compiled from such system approaches will reinforce the criteria for the selection of LUT-adapted fungal strains for evaluation as biological control agents. Future biological control strategies may not only include strains that are most virulent to particular pest insects but also those adapted to specific habitats, crops, or even sites. Such strategies may also include the manipulation of LUTs, for instance, by the cultivation of plant species that interact with these beneficial fungal species and support their growth, even in the occasional absence of a particular pest.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/microorganisms9071380/s1, Figure S1: Georeferenced sampling sites of the three land-use types, comprising 10 arable land sites, 10 grassland site, and 10 forest sites, Table S1: Summary information of samples, Table S2: Site metadata information: description of environmental factors of each site, Table S3: Means and ANOVA of Metarhizium spp. colony-forming units (CFU) g−1 of soil dry weight in each site. Affiliation of each site to one of the three abundance groups based on mean CFU is indicated, Table S4: Summary of 15 soil and environmental factors affecting the abundance groups observed among the 30 sites: high, medium, and low Metarhizium abundance (CFU g−1 of soil dry weight), Table S5: Summary of 15 soil and environmental factors affecting two Metarhizium abundance (CFU g−1 of soil dry weight) groups, medium and low, in arable land, Table S6: Summary of 15 soil and environmental factors affecting two Metarhizium abundance (CFU g−1 of soil dry weight) groups, high and medium, in grassland, Table S7: Summary of 15 soil and environmental factors affecting two Metarhizium abundance (CFU g−s1 of soil dry weight) groups, medium and low, in forest, Table S8: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. brunneum population structures among Microorganisms 2021, 9, 1380 15 of 19

three land-use types. Analyses were based on Bray–Curtis distances, Table S9: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. brunneum population structures in arable land and grassland. Analyses were based on Bray-Curtis distances, Table S10: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. brunneum population structures within arable land. Analyses were based on Bray–Curtis distances, Table S11: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. brunneum population structures within grassland. Analyses were based on Bray–Curtis distances, Table S12: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. brunneum population structures within forest sites. Analyses were based on Bray–Curtis distances, Table S13: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. robertsii population structures among three land-use types. Analyses were based on Bray–Curtis distances, Table S14: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. robertsii population structures within arable land. Analyses were based on Bray– Curtis distances, Table S15: Overall permutational multivariate analysis of variance (PERMANOVA) for individual soil and environmental factors that significantly affect M. robertsii population structure within grassland. Analyses were based on Bray–Curtis distances, Table S16: Summary of environmental factors at the ten sites of each land-use type. Author Contributions: Wrote the paper: M.F.-B. and J.E.; reviewed paper: M.F.-B., J.E., F.G., J.M., F.W., and A.H.; analyzed data: M.F.-B., F.G., and J.M.; performed experiments: M.F.-B., A.H., and J.E.; critically discussed results: M.F.-B., F.G., J.M., J.E., and F.W.; designed experiments: J.E. and F.W.; supervision: J.E.; funding acquisition: J.E. and M.F.-B. All authors have read and agreed to the published version of the manuscript. Funding: This project has received funding from the European Union’s Horizon 2020 research and innovation program under Marie Skłodowska-Curie grant agreement no. 794526. Data Availability Statement: All additional data can be obtained from the corresponding author upon reasonable request. Acknowledgments: We wish to thank Tabea Koch, Sonja Reinhard, and Stephanie Pfister for technical help and the NABO team at Agroscope for the collection of soil samples. Conflicts of Interest: The authors declare no conflict of interest.

Appendix A

Table A1. GenBank accession numbers of the EF1a sequences used to construct the phylogenetic tree in Figure3 of one representative isolate per MLG.

GeneBank Isolate MLG Species Accession Number MZ297396 95-III-2 MLG12 M. robertsii MZ297397 87-III-4 MLG19 M. brunneum MZ297398 77-III-5 MLG16 M. brunneum MZ297399 77-I-1 MLG20 M. brunneum MZ297400 70-III-1 MLG15 M. brunneum MZ297401 70-II-1 MLG7 M. guizhouense MZ297402 54-I-4 MLG1 M. robertsii MZ297403 45-III-6 MLG4 M. brunneum MZ297404 45-I-1 MLG6 M. brunneum MZ297405 35-III-3 MLG8 M. brunneum MZ297406 35-II-5 MLG9 M. brunneum Microorganisms 2021, 9, 1380 16 of 19

Table A1. Cont.

GeneBank Isolate MLG Species Accession Number MZ297407 35-I-6 MLG17 M. brunneum MZ297408 33-I-6 MLG2 M. robertsii MZ297409 33-I-5 MLG3 M. robertsii MZ297410 30-II-6 MLG21 M. guizhouense MZ297411 30-I-3 MLG18 M. brunneum MZ297412 28-II-2 MLG22 M. guizhouense MZ297413 28-I-1 MLG11 M. robertsii MZ297414 18-III-3 MLG5 M. brunneum MZ297415 7-III-6 MLG10 M. brunneum MZ297416 6-I-2 MLG14 M. brunneum MZ297417 1-III-4 MLG13 M. robertsii

References 1. Brunner-Mendoza, C.; Reyes-Montes, M.R.; Moonjely, S.; Bidochka, M.J.; Toriello, C. A review on the genus Metarhizium as an entomopathogenic microbial biocontrol agent with emphasis on its use and utility in Mexico. Biocontrol. Sci. Technol. 2019, 29, 83–102. [CrossRef] 2. Goettel, M.S.; Eilenberg, J.; Glare, T. Entomopathogenic fungi and their role in regulation of insect populations. In Comprehensive Molecular Insect Science; Gilbert, L.I., Gill, S.S., Eds.; Elsevier: Boston, MA, USA, 2005; pp. 361–406. 3. Lawo, N.C.; Mahon, R.J.; Milner, R.J.; Sarmah, B.K.; Higgins, T.J.V.; Romeis, J. Effectiveness of bacillus thuringiensis-transgenic chickpeas and the entomopathogenic fungus in controlling Helicoverpa armigera (: Noctuidae). mBio 2008, 74, 4381–4389. [CrossRef] 4. Pilz, C.; Wegensteiner, R.; Keller, S. Selection of entomopathogenic fungi for the control of the western corn rootworm Diabrotica virgifera virgifera. J. Appl. Entomol. 2007, 131, 426–431. [CrossRef] 5. Zimmermann, G. Review on safety of the entomopathogenic fungus Metarhizium anisopliae. Biocontrol. Sci. Technol. 2007, 17, 879–920. [CrossRef] 6. Maniania, N.K.; Ekesi, S. The use of entomopathogenic fungi in the control of tsetse flies. J. Invertebr. Pathol. 2013, 112, 83–88. [CrossRef][PubMed] 7. Brandl, M.A.; Schumann, M.; Przyklenk, M.; Patel, A.; Vidal, S. Wireworm damage reduction in potatoes with an attract-and-kill strategy using Metarhizium brunneum. J. Pest. Sci. 2016, 90, 479–493. [CrossRef] 8. Faria, M.; Wraight, S.P. Mycoinsecticides and Mycoacaricides: A comprehensive list with worldwide coverage and international classification of formulation types. Biol. Control. 2007, 43, 237–256. [CrossRef] 9. Vega, F.E.; Goettel, M.S.; Blackwell, M.; Chandler, D.; Jackson, M.A.; Keller, S.; Koike, M.; Maniania, N.K.; Monzon, A.; Ownley, B.H.; et al. Fungal entomopathogens: New insights on their ecology. Fungal Ecol. 2009, 2, 149–159. [CrossRef] 10. Hu, G.; St. Leger, J. Field studies using a recombinant mycoinsecticide (Metarhizium anisopliae) reveal that it is rhizosphere competent. Appl. Environ. Microbiol. 2002, 68, 6383–6387. [CrossRef] 11. Behie, S.W.; Zelisko, P.M.; Bidochka, M.J. Ubiquity of insect-derived nitrogen transfer to plants by endophytic insect-pathogenic fungi: An additional branch of the soil nitrogen cycle. Appl. Environ. Microbiol. 2012, 80, 1553–1560. [CrossRef] 12. Barelli, L.; Moonjely, S.; Behie, S.W.; Bidochka, M.J. Fungi with multifunctional lifestyles: Endophytic insect pathogenic fungi. Plant. Mol. Biol. 2016, 90, 657–664. [CrossRef] 13. Raya-Diaz, S.; Sanchez-Rodriguez, A.R.; Segura-Fernandez, J.M.; del Campillo, M.C.; Quesada-Moraga, E. Entomopathogenic fungi-based mechanisms for improved Fe nutrition in sorghum plants grown on calcareous substrates. PLoS ONE 2017, 12, e0185903. [CrossRef] 14. Jaber, L.R.; Enkerli, J. Fungal entomopathogens as endophytes: Can they promote plant growth? Biocontrol. Sci. Technol. 2017, 27, 28–41. [CrossRef] 15. Rangel, D.E.N.; Alder-Rangel, A.; Dadachova, E.; Finlay, R.D.; Dijksterhuis, M.K.J.; Braga, G.U.L.; Corrochano, L.M.; Hallsworth, J.E. Fungal stress biology: A preface to the fungal stress responses special edition. Curr. Genet. 2015, 61, 231–238. [CrossRef] 16. Khan, A.L.; Hussain, J.; Al-Harrasi, A.; Al-Rawahi, A.; Lee, I.J. Endophytic fungi: Resource for gibberellins and crop abiotic stress resistance. Crit. Rev. Biotechnol. 2013, 35, 62–74. [CrossRef][PubMed] 17. Meyling, N.V.; Eilenberg, J. Ecology of the entomopathogenic fungi Beauveria bassiana and Metarhizium anisopliae in temperate agroecosystems: Potential for conservation biological control. Biol. Control. 2007, 43, 145–155. [CrossRef] 18. Foley, J.A.; Defries, R.S.; Asner, G.P. Global consequences of land use. Science. 2005, 309. [CrossRef] 19. Vänninen, I. Distribution and occurrence of four entomopathogenic fungi in Finland: Effect of geographical location, habitat type and soil type. Mycol. Res. 1996, 100, 93–101. [CrossRef] 20. Jabbour, R.; Barbercheck, M.E. Soil management effects on entomopathogenic fungi during the transition to organic agriculture in a feed grain rotation. Biol. Control. 2009, 51, 435–443. [CrossRef] Microorganisms 2021, 9, 1380 17 of 19

21. Kepler, R.M.; Ugine, T.A.; Maul, J.E.; Cavigelli, M.A.; Rehner, S.A. Community composition and population genetics of insect pathogenic fungi in the genus Metarhizium from soils of a long-term agricultural research system. Environ. Microbiol. 2015, 8, 2791–2804. [CrossRef] 22. Clifton, E.H.; Jaronski, S.T.; Hodgson, E.W.; Gassmann, A.J. Abundance of soil-borne entomopathogenic fungi in organic and conventional fields in the mid-western USA with an emphasis on the effect of herbicides and fungicides on fungal persistence. Biol. Control. 2015, 51, 435–443. [CrossRef] 23. Bidochka, M.J.; Kasperski, J.E.; Wild, G.A.M. Occurrence of the entomopathogenic fungi Metarhizium anisopliae and Beauveria bassiana in soils from temperate and near-northern habitats. Canad. J. Bot. 1998, 76, 1198–1204. [CrossRef] 24. Quesada-Moraga, E.; Navas-Cortés, J.A.; Maranhao, E.A.A.; Ortiz-Urquiza, A.; Santiago-Álvarez, C. Factors affecting the occurrence and distribution of entomopathogenic fungi in natural and cultivated soils. Mycol. Res. 2007, 111, 947– 966. [CrossRef][PubMed] 25. Garrido-Jurado, I.; Fernández-Bravo, M.; Campos, C.; Quesada-Moraga, E. Diversity of entomopathogenic Hypocreales in soil and phylloplane of five Mediterranean cropping systems. J. Invertebr. Pathol. 2015, 130, 97–106. [CrossRef] 26. Nielsen, C.; Skovgard, H.; Steenberg, T. Effect of Metarhizium anisopliae (Deuteromycotina: ) on survival and repro- duction of the filth fly parasitoid, cameroni (: ). Environ. Entomol. 2004, 34, 133–139. [CrossRef] 27. Keller, S.; Kessler, P.; Schweizer, C. Distribution of insect pathogenic soil fungi in Switzerland with special reference to Beauveria brongniartii and Metharhizium anisopliae. BioControl. 2003, 48, 307–319. [CrossRef] 28. Schneider, S.; Widmer, F.; Jacot, K.; Kolliker, R.; Enkerli, J. Spatial distribution of Metarhizium clade 1 in agricultural landscapes with arable land and different semi-natural habitats. Appl. Soil. Ecol. 2012, 52, 20–28. [CrossRef] 29. Klingen, I.; Eilenberg, J.; Meadow, R. Effects of farming system, field margins and bait insect on the occurrence of insect pathogenic fungi in soils. Agric. Ecosyst. Environ. 2002, 91, 191–198. [CrossRef] 30. Mi˛etkiewski,R.T.; Pell, K.J.; Clark, S.J. Influence of pesticide use on the natural occurrence of entomopathogenic fungi in arable soil in the UK: Field and laboratory comparisons. Biocontrol. Sci. Technol. 1997, 7, 565–575. [CrossRef] 31. Cabrera-Mora, J.A.; Guzmán-Franco, A.W.; Santillán-Galicia, M.T.; Tamayo-Mejía, F. Niche separation of species of ento- mopathogenic fungi within the genera Metarhizium and Beauveria in different cropping systems in Mexico. Fungal Ecol. 2019, 39, 349–355. [CrossRef] 32. Wyrebek, M.; Huber, C.; Sasan, R.K.; Bidochka, M.J. Three sympatrically occurring species of Metarhizium show plant rhizosphere specificity. Microbiology 2011, 157, 2904–2911. [CrossRef] 33. Bidochka, M.J.; Kamp, A.M.; Lavender, T.M.; Dekoning, J.; De Croos, J.N.A. Habitat association in two genetic groups of the insect-pathogenic fungus Metarhizium anisopliae: Uncovering cryptic species? Appl. Environ. Microbiol. 2001, 67, 1335–1342. [CrossRef][PubMed] 34. Bidochka, M.J.; Small, C.L. Phylogeography of Metarhizium, an insect pathogenic fungus. In Insect-Fungal Associations: Ecology and Evolution; Vega, F.E., Blackwell, M., Eds.; Oxford University Press: New York, NY, USA, 2005; pp. 75–118. 35. Bischoff, J.F.; Rehner, S.A.; Humber, R.A. A multilocus phylogeny of the Metarhizium anisopliae lineage. Mycologia. 2009, 101, 512–530. [CrossRef][PubMed] 36. Steinwender, B.M.; Enkerli, J.; Widmer, F.; Eilenberg, J.; Thorup-Kristensen, K.; Meyling, N.V. Molecular diversity of the ento- mopathogenic fungal Metarhizium community within an agroecosystem. J. Invertebr. Pathol. 2014, 123, 6–12. [CrossRef][PubMed] 37. Castro, T.; Mayerhofer, J.; Enkerli, J.; Eilenberg, J.; Meyling, N.V.; Moral, R.D.; Demetrio, C.G.B.; Delalibera, I. Persistence of Brazilian isolates of the entomopathogenic fungi Metarhizium anisopliae and M. robertsii in strawberry crop soil after soil drench application. Agric. Ecosyst. Environ. 2016, 233, 361–369. [CrossRef] 38. Inglis, G.D.; Duke, G.M.; Goettel, M.S.; Kabaluk, J.T.; Ortega-Polo, R. Biogeography and genotypic diversity of Metarhizium brunneum and Metarhizium robertsii in northwestern North American soils. Can. J. Microbiol. 2019, 65, 261–281. [CrossRef] 39. Ekesi, S.; Maniania, N.K.; Ampong-Nyarko, K. Effect of temperature on germination, radial growth and virulence of Metarhizium anisopliae and Beauveria bassiana on Megalurothrips sjostedti. Biocontrol Sci. Technol. 1999, 9, 177–185. [CrossRef] 40. Rangel, D.E.N.; Braga, G.U.L.; Anderson, A.J.; Roberts, D.W. Variability in conidial thermotolerance of Metarhizium anisopliae isolates from different geographic regions. J. Invertebr. Pathol. 2005, 88, 116–125. [CrossRef][PubMed] 41. Fernández-Bravo, M.; Flores-León, A.; Calero-López, S.; Gutiérrez-Sánchez, F.; Valverde-García, P.; Quesada-Moraga, E. UV-B radiation-related effects on conidial inactivation and virulence against Ceratitis capitata (Wiedemann) (Diptera; Tephritidae) of phylloplane and soil Metarhizium sp. Strains. J. Invertebr. Pathol. 2017, 148, 142–151. [CrossRef] 42. McGuire, A.V.; Northfield, T.D. Tropical Occurrence and Agricultural Importance of Beauveria bassiana and Metarhizium anisopliae. Front. Sustain. Food. Syst. 2020, 4, 6. [CrossRef] 43. Uzman, D.; Pliestera, J.; Leyerb, I.; Entlingc, M.H.; Reinekea, A. Drivers of entomopathogenic fungi presence in organic and conventional vineyard soils. Appl. Soil Ecol. 2019, 133, 89–97. [CrossRef] 44. Medo, J.; Cagán, M. Factors affecting the occurrence of entomopathogenic fungi in soils of Slovakia as revealed using two methods. Biol. Control. 2011, 59, 200–208. [CrossRef] 45. Gubler, A.; Schwab, P.; Wächter, D.; Meuli, R.G.; Keller, A. Ergebnisse der Nationalen Bodenbeobachtung (NABO) 1985–2009. Zustand und Veränderungen der anorganischen Schadstoffe und Bodenbegleitparameter; Bundesamt für Umwelt—Umwelt-Zustand: Bern, Germany, 2015; pp. 1–81. Microorganisms 2021, 9, 1380 18 of 19

46. Gschwend, F.; Hartmann, M.; Hug, A.; Enkerli, J.; Gubler, A.; Frey, B.; Meuli, R.G.; Widmer, F. Long-term stability of soil bacterial and fungal community structures revealed in their abundant and rare fractions. Mol. Ecol. 2021, in press. [CrossRef][PubMed] 47. Gubler, A.; Wächter, D.; Schwab, P.; Müller, M.; Keller, A. Twenty-five years of observations of soil organic carbon in Swiss croplands showing stability overall but with some divergent trends. Environ. Monit. Assess. 2019, 191, 277. [CrossRef] 48. Strasser, H.; Forer, A.; Schinner, F. Development of media for the selective isolation and maintenance of virulence of Beauveria brongniartii. In Microbial Control of Soil Dwelling Pests; Jackson, T., Glare, T., Eds.; Christchurch, AgResearch Publisher: Lincoln, New Zealand, 1996; pp. 125–130. 49. Barnett, H.L.; Hunter, B.B. Illustrated Genera of Imperfect Fungi; APS Press: St. Paul, MN, USA, 2006. 50. Rehner, S.A.; Buckley, E.P. A Beauveria phylogeny inferred from nuclear ITS and EF1-alpha sequences: Evidence for cryptic diversification and links to Cordyceps teleomorphs. Mycologia 2005, 97, 84–98. [CrossRef] 51. Mayerhofer, J.; Rauch, H.; Hartmann, M.; Widmer, F.; Gschwend, F.; Strasser, H.; Leuchtmann, A.; Enkerli, J. Response of soil microbial communities to the application of a formulated Metarhizium brunneum biocontrol strain. Biocontrol Sci. Technol. 2019, 29, 547–564. [CrossRef] 52. Kumar, S.; Stecher, G.; Li, M.; Knyaz, C.; Tamura, K. MEGA X: Molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol. 2018, 35, 1547–1549. [CrossRef] 53. Kimura, M. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 1980, 16, 111–120. [CrossRef] 54. Mayerhofer, J.; Lutz, A.; Widmer, F.; Rehner, S.A.; Leuchtmann, A.; Enkerli, J. Multiplexed microsatellite markers for seven Metarhizium species. J. Invertebr. Pathol. 2015, 132, 132–134. [CrossRef] 55. Kamvar, Z.N.; Tabima, J.F.; Grünwald, N.J. Poppr: An R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2014, 2, e281. [CrossRef][PubMed] 56. Shannon, C.E. A mathematical theory of communication. Bell Syst. Tech. 1948, 27, 379–423. [CrossRef] 57. Pielou, E. Ecological Diversity; John Wiley and Sons: Hoboken, NJ, USA, 1975. 58. Ludwig, J.; Reynolds, J. Statistical Ecology. A Primer on Methods and Computing; John Wiley and Sons: Hoboken, NY, USA, 1988. 59. Grünwald, N.J.; Goodwin, S.B.; Milgroom, M.G.; Fry, W.E. Analysis of genotypic diversity data for populations of microorganisms. Phytopathology 2003, 93, 738–746. [CrossRef][PubMed] 60. Bray, J.R.; Curtis, J.T. An ordination of the upland forest communities of southern Wisconsin. Ecol. Monogr. 1957, 27, 326–349. [CrossRef] 61. Hervé, M. RVAideMemoire: Testing and Plotting Procedures for Biostatistics. R Package Version 0.9-69-3; RStudoi: Boston, MA, USA, 2018. 62. Oksanen, J.; Blanchet, G.F.; Friendly, M.; Kindt, R.; Legendre, P.; McGlinn, D.; Wagner, H. Vegan: Community Ecology Package. R Package Version 2.4-6; RStudoi: Boston, MA, USA, 2018; Available online: http://CRAN.R-project.org/package=vegan (accessed on 27 May 2021). 63. R-Core-Team. R: A Language and Environment For Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2016. 64. Gower, J.C. Some distance properties of latent root and vector methods used in multivariate analysis. Biometrika 1966, 53, 325–328. [CrossRef] 65. Shah, F.A.; Wang, C.S.; Butt, T.M. Nutrition influence growth and virulence of the insect-pathogenic fungus Metarhizium anisopliae. FEMS Microbiol. Lett. 2005, 251, 259–266. [CrossRef] 66. Gao, L.; Sun, M.H.; Liu, X.Z.; Che, Y.S. Effects of carbon concentration and carbon to nitrogen ratio on the growth and sporulation of several biocontrol fungi. Mycol. Res. 2007, 111, 87–92. [CrossRef] 67. Ortiz-Urquiza, A.; Garrido-Jurado, I.; Borrego, A.; Quesada-Moraga, E. Effects of cultural conditions on fungal biomass, blastospore yields and toxicity of fungal secreted proteins in batch cultures of Metarhizium anisopliae (: Hypocreales). Pest. Manag. Sci. 2010, 66, 725–735. [CrossRef] 68. Sehler, R.; Li, J.; Reager, J.T.; Ye, H. Investigating relationship between soil moisture and precipitation globally using remote sensing observations. J. Contemp. Water Res. Educ. 2019, 168, 106–118. [CrossRef] 69. Garrido-Jurado, I.; Valverde-Garcia, P.; Quesada-Moraga, E. Use of a multiple logistic regression model to determine the effects of soil moisture and temperature on the virulence of entomopathogenic fungi against pre-imaginal Mediterranean fruit fly Ceratitis capitata. Biol. Control. 2011, 59, 366–372. [CrossRef] 70. Sosa-Gómez, D.R.; Delpin, K.E.; Moscardi, F.; Farias, J.R.B. Natural occurrence of the entomopathogenic fungi Metarhizium, Beauveria and Paecilomyces in soybean under till and no-till cultivation systems. Neotrop. Entomol. 2001, 30, 407–410. [CrossRef] 71. Hummel, R.L.; Walgenbach, J.F.; Barbercheck, M.E.; Kennedy, G.G.; Hoyt, G.D.; Arellano, C. Effects of production practices on soil-borne entomopathogens in western North Carolina vegetable systems. Environ. Entomol. 2002, 31, 84–91. [CrossRef] 72. Inglis, G.D.; Duke, G.M.; Goettel, M.S.; Kabaluk, J.T. Genetic diversity of Metarhizium anisopliae var. anisopliae in southwestern British Columbia. J. Invertebr. Pathol. 2007, 98, 101–113. [CrossRef] 73. Rehner, S.A. Genetic structure of Metarhizium species in western USA: Finite populations composed of divergent clonal lineages with limited evidence for recent recombination. J. Invertebr. Pathol. 2020, 177, 107491. [CrossRef][PubMed] Microorganisms 2021, 9, 1380 19 of 19

74. Fisher, J.J.; Rehner, S.A.; Bruck, D.J. Diversity of rhizosphere associated entomopathogenic fungi of perennial herbs, shrubs and coniferous trees. J. Invertebr. Pathol. 2011, 106, 289–295. [CrossRef][PubMed] 75. Arribas, P.; Andújar, C.; Salces-Castellano, A.; Emerson, B.C.; Vogler, A.P. The limited spatial scale of dispersal in soil arthropods revealed with whole-community haplotype-level metabarcoding. Mol. Ecol. 2021, 30, 48–61. [CrossRef] 76. Dromph, K.M. Collembolans as vectors of entomopathogenic fungi. Pedobiologia 2003, 47, 245–256. [CrossRef]